[Author’s note: I am absolutely confident I got this definition from somewhere else, but I’ve looked for it extensively and haven’t been able to find it, and I’d like to be able to reference it, so I’m writing it up. Sorry, original inventor of this definition, whom I have failed to credit.]
Definitions of effective altruism are often very vague. The Centre for Effective Altruism defines effective altruism as “using evidence and reason to figure out how to benefit others as much as possible, and taking action on that basis.” The Effective Altruism Foundation defines effective altruism as “a philosophy and a social movement which holds that actively helping others is of central moral importance, and approaches the choice of possible strategies in a rational and scientific way.” Various book titles define it as Doing Good Better or The Most Good You Can Do.
I’ve talked to people who are confused by these definitions. It seems like in principle feeding hungry people in America or sheltering stray cats could be done in a rational and scientific way, and they certainly involves actively helping others; some people are confused that effective altruists don’t tell them how to most effectively help stray cats or poor Americans. Other people believe that effective altruism is solely about donating to global health charities that have been shown to work in randomized controlled trials and are confused by the fact that many effective altruists are involved in causes with scanty or no peer-reviewed scientific backing.
I’d like to suggest a better definition. In many ways, effective altruism is a big tent: it is clearly not limited to a single cause or way of arguing. However, I think there is a distinctive effective altruism approach to doing good, which is worth defining.
Effective altruism is secular. It does not recommend charities that most effectively get people into Heaven, cause people to have a personal relationship with Jesus, or cause people to reach enlightenment, despite the many religious people that believe that these are more important ways to benefit others than preventing nuclear war or eradicating malaria.
Effective altruism is outcome-oriented. When you justify a course of action from an effective altruist point of view, you must explain why you believe this course of action will cause some specific good thing to go up or some specific bad thing to go down– how it will reduce deaths, make people healthier, improve education, prevent human extinction, or cause fewer animals to suffer torturous lives. You cannot justify the course of action by saying that it is what a virtuous person would do, or that there is some rule that says that everyone should do it, or that it respects human dignity, or that it seems like the sort of thing someone ought to be doing even if it has no effects whatsoever.
Effective altruism is maximizing. It is, as the book title says, about doing the most good you can do. An effective altruist approach does not involve listing off twenty things that we think pass a certain threshold for goodness. It involves saying what you think the single best thing is. Of course, there is uncertainty, so you might say “I don’t know which of these five things are best.” And some things might be better for one person than for another. If trying to be vegan will put you in the hospital, then obviously being vegan is not The Most Good You Can Do. If you’re an amazing political activist and a mediocre AI alignment researcher, then it might be best for you to be a political activist, even if it’s best for an amazing AI alignment researcher to be an AI alignment researcher.
Effective altruism is cause-impartial. Many people choose which cause they work on for reasons other than trying to have the most positive effect on the world. They choose a cause that they’re passionate about, or that they have a personal connection to, or that makes them feel warm and fuzzy feelings. They donate to a particular charity out of habit or because someone asked them to. Effective altruism, however, is impartial between causes: effective altruism recommends that people do whatever seems to be best, not whatever gives them the warmest and fuzziest feelings.
Effective altruism is egalitarian. Effective altruism values all people equally: that is, from the effective altruism perspective, saving the life of a baby in Africa is exactly as good as saving the life of a baby in America, which is exactly as good as saving the life of Ozy’s baby specifically. Effective altruism does not value some beings more and other beings less because they live in different places, or because one is cuter or more sympathetic than the other, or because one has a different skin tone than another. Despite some difficulties about how to justify it philosophically, effective altruism generally believes that future people are as important as present people. While there is controversy about to what extent nonhuman animals should be valued, effective altruism is not speciesist: if effective altruism should not value nonhuman animals, it is because they can’t feel pain, can’t suffer, are not conscious, or are incapable of having long-term preferences, not because of their species membership.
I have been careful throughout this post to say “effective altruism” rather than “effective altruists.” Certainly, it’s not an accident that effective altruists are typically atheists with consequentialist ethical systems. There’s a natural harmony between secular, outcome-oriented effective altruism and atheist consequentialism. If a person believes that converting people is so important that it’s a waste of time to feed bellies instead of souls, or that outcomes don’t matter at all, or that people in America are millions of times more important than people in Africa, they are unlikely to benefit much from effective altruism and can safely ignore our recommendations. But many people who do not fully agree with all the values of effective altruism can derive value from an effective altruism approach.
For example, you might rule out certain courses of action like killing people, even if they lead to the best outcomes. However, since Assassins Without Borders is not very likely to be recommended as a top charity any time soon, an outcome-oriented approach can still help you figure out your career and donation decisions. You might be religious but believe in an obligation to help people in this world rather than just the next; perhaps you split your effort and donations between religious and secular causes, and use effective altruism to guide your secular effort.
In fact, very very few effective altruists apply the effective altruism approach to every aspect of their lives. Professional effective altruists write a lot of fanfiction, which would be weird if they were trying to maximize the amount of good they’re doing with every moment. And I for one am willing to spend much more money to save the life of my baby than I am to save the life of some unrelated baby. Nevertheless, I find effective altruism very valuable in figuring out how to do the most good with the time, effort, and energy I’m willing to spend on that.
Relevant to this, in the following article MacAskill provides the following account of what EA is:
Also, you've accidentally posted the same thing three times, if you hadn't noticed already.
I like "science-aligned" better than "secular", since the former implies the latter as well as a bunch of other important concepts.
Also, it's worth noting that "everyone's welfare is to count equally" in Will's account is approximately equivalent to "effective altruism values all people equally" in Ozymandias' account, but neither of them imply the following paraphrase: "from the effective altruism perspective, saving the life of a baby in Africa is exactly as good as saving the life of a baby in America, which is exactly as good as saving the life of Ozy’s baby specifically." I understand the intention of that phrase, but actually I'd save whichever baby would grow up to have the best life. Is there any better concrete description of what impartiality actually implies?
"Whichever will have the best life" seems very compatible with "welfare." I agree there are a lot of considerations that aren't obviously indicated by any of these definitions, though, even if they're compatible.
Yes, I'd be excited to always include something about epistemics, such as scientific mindset. One can then argue about evidence instead of whether something qualifies as secular, which seems only relevant insofar as it is a weak predictor of well-evidenced-ness. In particular, while I don't assign it high credence, I would not be hasty to rule out maximising 'enlightenment' as one of my end goals. Terminal goals are weird.
Notably, without an epistemic/scientific part to the definition, it is unclear how to distinguish many current EA approaches from e.g. Playpumps, which is a secular project which was hyped for its extraordinary outcomes in helping people far away. Looking forward, I also think that strong epistemics are how long-term-focused EA efforts can continue to be more useful than regular futurism.